There are many active research projects accessing and applying shared ADNI data. Use the search above to find specific research focuses on the active ADNI investigations. This information is requested annually as a requirement for data access.
| Principal Investigator | |
| Principal Investigator's Name: | Feiyan Feng |
| Institution: | Shandong Normal University |
| Department: | information science and engineering |
| Country: | |
| Proposed Analysis: | The aim of our study is to integrate genetic, imaging, and behavioral data from different scales and modalities to investigate the underlying mechanisms of Alzheimer's disease (AD). AD is a common neurodegenerative disease that primarily affects the elderly population. It is characterized by an irreversible disease progression, and its precise pathogenic mechanisms are still unclear, with a lack of effective treatment options. By comprehensively analyzing data from multiple sources, we hope to gain a comprehensive understanding of the pathogenic mechanisms of AD, enabling early diagnosis and timely intervention to slow down disease progression. In this study, we will employ the following analysis methods: 1. Data Integration and Standardization: We will collect various data types, including genetic, imaging, and behavioral data, from different sources. These data will be integrated and standardized to ensure data consistency and comparability. 2. Imaging Data Analysis: Utilizing neuroimaging techniques such as Magnetic Resonance Imaging (MRI) and Positron Emission Tomography (PET), we will compare and analyze AD patients with a healthy control group to reveal AD pathological features and changes in brain structure and function. 3. Addressing the High-Dimensional Small Sample Problem: To tackle the high-dimensional small sample problem in brain imaging genetics data, we plan to employ an ensemble learning approach. This method will integrate multimodal data into binary gene-brain region association features using a correlation representation approach. Additionally, ensemble pruning methods will be applied to extract disease-discriminative gene-brain region association features for AD disease classification. This approach outperforms other methods in terms of classification performance and feature selection, providing strong support for clinical diagnosis and identification of pathogenic mechanisms in AD. 4. Addressing the Heterogeneity of Multimodal Data: To address the heterogeneity issue in multimodal data, we will utilize graph capsule convolutional networks. By constructing a disease-relevant graph using genes and brain regions as nodes, we will integrate different modalities of data. Leveraging disentanglement techniques, we will project the heterogeneous information of multimodal data into a set of latent components to construct graph capsules. The graph capsule convolutional network will then capture disease-specific information flow between pathogenic factors. Experimental results have demonstrated the advanced capabilities of this method in predicting progressive mild cognitive impairment and identifying AD-related pathogenic factors. Through the comprehensive application of the aforementioned analysis methods, we aim to gain a comprehensive understanding of the pathogenic mechanisms of AD, providing scientific evidence for early diagnosis and intervention and driving further advancements in the treatment and management of AD. Your dataset is crucial for our research, and we sincerely appreciate your consideration of our application. |
| Additional Investigators |
